Prolific posting day for me. I saw this on Twitter, An academic paper about system over estimate of distance by GPS devices.
There is a lot of heavy duty math in the paper and if I manage to understand it, it will take a while. But the basic gist is that the larger the error in the estimated positions along the route trajectory, the more cumulative distance errors occurs.
In the pedestrian example they provided, the error peaked at 2.5 meters of over estimation at 45 minutes into the experiment.
Not all that much error, but it gives you something to think about.